Words matter. In law, they are everything. They are the medium through which meaning is conveyed, disputes are resolved, and justice is done, or undone. They define obligations, impose duties, confer rights, and, when ill-chosen, they sow confusion. In a sense, my entire working life, has been devoted to words, in one form or another.
The tragic case of Derek Bentley demonstrates the fatal consequences that can arise from ambiguity in language. In 1953, Bentley, a young man of limited capacity, was convicted of murder and hanged. During a police confrontation on a rooftop in Croydon, he uttered the phrase “Let him have it, Chris” to his armed accomplice, the 16 year old Chris Craig, who had brought a revolver with him on that fateful evening.
The meaning of this phrase was pivotal. Did Bentley mean for Chris Craig to surrender the weapon? Or was he inciting him to shoot? The jury, it seems, interpreted it as the latter. Forty years later, Bentley received a posthumous pardon. In 1998, the Court of Appeal quashed his conviction, noting a misdirection by the trial judge. But by then, Bentley was long dead. The ambiguity of that one sentence cost a young man his life.
If ambiguity in spoken words can hang a man, ambiguity in written words can ruin a legal compromise. In the context of litigation, the words used in correspondence and settlement can have lasting and expensive consequences. One common flashpoint arises when parties believe they have reached a settlement of the substantive issues but remain at odds as to whether, and how, costs are to be dealt with.
It is not uncommon for parties to exchange terms of settlement with no mention of costs. Or for one party to proceed on the basis that costs will follow the event, while the other believes that each side will bear its own. Or worse still, for one party to accept a settlement offer with an added term regarding costs that was never contemplated by the offeror. In such cases, the parties may find themselves arguing not only about costs but also about whether a concluded contract exists at all.
Under orthodox contract law principles, there must be offer and acceptance, supported by consideration and characterised by sufficient certainty. The mirror image rule governs an acceptance: it must match the offer in every material respect. If it does not, the purported acceptance is a counter-offer, which extinguishes the original offer. . The question in costs settlements is whether a response that varies the costs position amounts to a true acceptance or a counter-offer. If the terms on costs are essential—and in litigation, they usually are—then divergence may mean no binding agreement was ever formed.
Assuming there is a binding agreement, the next question is what the agreement means. The leading authority remains Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 WLR 896, in which Lord Hoffmann famously set out five principles that govern contractual interpretation.
First, the meaning of a contract is what the document would convey to a reasonable person with all the background knowledge reasonably available to the parties at the time.
Second, the background includes absolutely anything which would have affected the way in which the language of the document would have been understood by a reasonable person.
Third, the law excludes from the admissible background the previous negotiations of the parties and their declarations of subjective intent.
Fourth, the meaning of the words is not the same as their literal meaning; context is key.
Fifth, the court must avoid interpretations that produce irrational or absurd results.
The approach in ICS prioritises objective interpretation. What the parties meant to say is not the issue; what matters is what they actually said, when viewed in its proper context. This context, the so-called “matrix of fact”, must be assessed as it stood at the time of agreement, not with the benefit of hindsight. The words used are to be read as a whole and given the meaning a reasonable person would attribute to them. A court will not rewrite a contract simply because one party is now dissatisfied with the result.
The tension between textualism and contextualism in contractual interpretation has occupied the appellate courts in recent years. In Rainy Sky SA v Kookmin Bank [2011] UKSC 50; [2011] 1 WLR 2900, the Supreme Court favoured a purposive approach, affirming that where a contractual term is open to more than one construction, the court should adopt the interpretation that is most consistent with business common sense. This pragmatic principle promotes commercial certainty and seeks to uphold what rational parties would have intended.
However, this approach met with apparent retrenchment in Arnold v Britton [2015] UKSC 36; [2015] AC 1619. There, the court placed greater emphasis on the actual language used, warning against a freewheeling search for business common sense that ignores the words chosen by the parties. Lord Neuberger, giving the leading judgment, stressed that commercial common sense should not be invoked to undervalue the importance of the language of the provision itself. Clarity of expression must be respected, even if it leads to an unwise or harsh outcome for one party.
A more reconciliatory note was struck in Wood v Capita Insurance Services Ltd [2017] UKSC 24; [2017] AC 1173. Lord Hodge held that textualism and contextualism are not conflicting approaches but tools in the interpretive toolkit. The extent to which each assists will depend on the nature of the contract, its sophistication, and the factual matrix. The judgment reiterated the importance of the language used but acknowledged that context can illuminate the meaning of even apparently straightforward terms.
This nuanced approach was further clarified in Lukoil Asia Pacific Pte Ltd v Ocean Tankers (Pte) Ltd [2018] EWHC 163 (Comm); [2018] 1 Lloyd’s Rep 654, where the court observed that textualism and contextualism are not competing ideologies but complementary methods. The ultimate aim remains the same: to identify the objective meaning of the words chosen by the parties.
But what if the parties have proceeded on a shared assumption about the effect of their settlement—say, that costs are to be borne in a certain way—even if that assumption is not supported by the written agreement? In such cases, the doctrine of estoppel by convention may apply.
The modern statement of this doctrine is found in Tinkler v HMRC [2021] UKSC 39; [2022] AC 886. The elements are as follows: (i) the parties must share a common assumption, (ii) they must act upon it in a way that affects their legal relations, (iii) there must be reliance by at least one party, (iv) that reliance must be detrimental, and (v) it must be unconscionable to allow the other party to depart from the shared assumption.
This is a demanding test. Estoppel by convention is not lightly established, and the burden of proof lies with the party seeking to rely on it. The assumption must be clear and unequivocal, the reliance must be causative, and the overall result must be such that departure from the assumption would work injustice.
In some cases, a settlement that appears to be concluded may be challenged on grounds that go beyond interpretation or estoppel. These include mistake, misrepresentation, and duress. A common mistake as to a material term may vitiate the agreement if it renders the contract void. Misrepresentation, whether fraudulent, negligent, or innocent, may render the contract voidable. Duress, including lawful act duress, may also justify rescission.
The concept of lawful act duress is relatively modern and was considered by the Supreme Court in Times Travel (UK) Ltd v Pakistan International Airlines Corporation [2021] UKSC 40. There, the court held that pressure amounting to duress must involve illegitimate conduct. While hard bargaining is not of itself unlawful, threats made with the intention of compelling agreement, in the absence of genuine belief in their legitimacy, may amount to duress. If one party threatens to allege fundamental dishonesty unless the other party agrees to forego costs, the court may well find the agreement vitiated by duress.
Particular care is needed when dealing with litigants in person. An unrepresented party may not appreciate the significance of omitting a costs clause. They may accept a settlement offer believing it includes costs, even if this is not stated. While the courts are slow to intervene in arm’s length agreements, they may scrutinise such settlements more closely to ensure there is genuine agreement, particularly where there is an obvious inequality of arms.
It bears repeating that ambiguity in settlement communications creates fertile ground for dispute. The very purpose of settlement is to bring finality. Yet when costs are left unaddressed or are dealt with in opaque or contradictory terms, that purpose is defeated. A poorly drafted settlement is an invitation to satellite litigation.
The prudent course is to state in express terms how costs are to be dealt with. If they are included, say so. If each side is to bear their own costs, say so. If costs are to be subject to assessment or agreed on a standard basis, specify the mechanism. In the absence of such clarity, the court may find there is no agreement at all—or that the agreement is susceptible to challenge on any of the grounds set out above.
Returning to Derek Bentley, one is struck by the power—and danger—of words. A single sentence, ambiguous and unclarified, ended a life. In civil litigation, words carry no such mortal burden, but they may determine outcomes of great financial and reputational weight. The law demands clarity, not perfection. But where clarity is lacking, the consequences can be severe.